CSS-1 Bitumen Emulsion: Specification, Tack Coat Use and Handling

CSS-1 is a cationic slow-setting bitumen emulsion, specified under ASTM D2397, Standard Specification for Cationic Emulsified Asphalt, and its AASHTO counterpart M208. Physically it is paving bitumen dispersed as fine droplets through a continuous water phase, held apart by an amine emulsifier that has been neutralised with acid so that every droplet carries a positive surface charge. What makes it a slow-setting grade is not the bitumen and not the water. It is the quantity and type of emulsifier: a CSS grade is stabilised heavily enough that the droplets survive contact with a mineral surface instead of collapsing onto it. That survival is the whole product, and every practical property of CSS-1 follows from it.

The comparison that makes the grade easiest to understand is with CRS-2, because the two sit in the same requirement table of the same standard and share exactly one property. Both are cationic, so both are drawn electrostatically to the negatively charged siliceous aggregate that most roads are built from, and both need mild or stainless steel equipment because both are acidic. After that they diverge completely: one is designed to break in seconds under a spread of chippings, the other is designed not to break while it is being mixed into a slurry or thinned to a quarter of its strength and sprayed at a fraction of a litre per square metre. There is no operation in which one substitutes for the other. The emulsion grades page tabulates the whole family; this page is about the slow-setting half of it.

Reading the designation, character by character

• C — cationic. The emulsifier is a fatty amine neutralised with acid, the droplets carry a positive charge, and the finished product is acidic. The leading letter also places the grade under ASTM D2397 rather than under ASTM D977, which covers the anionic family. It is proved on the certificate by the particle charge test to ASTM D244: direct current through two electrodes, and the bitumen must migrate to the cathode for the result to be reported positive. This is the only test that proves the family, and the difference between a cationic and an anionic product is invisible in the drum.
• SS — slow setting. The setting class states how quickly the grade is formulated to break, and therefore which operations it belongs to. Slow-setting is the most heavily stabilised class in either standard: it tolerates fine, high surface area material including Portland cement, and it is the only class that is routinely diluted with water. Its defining acceptance test is cement mixing to ASTM D6935, with a maximum of 2.0 % coagulum.
• 1 — the grade number. Across the emulsion naming system the digit is an index of body and binder content rather than a percentage or a quality ranking, so a 2 grade is more viscous and carries more residue than the corresponding 1 grade. There is a point specific to this family that is worth knowing: ASTM D2397 defines no CSS-2. The cationic slow-setting family consists of CSS-1 and CSS-1h and nothing else, and the same is true of SS-1 and SS-1h on the anionic side of ASTM D977. The digit is carried for consistency with the rest of the naming system rather than to distinguish CSS-1 from a heavier sibling that does not exist. An offer of CSS-2 against an ASTM specification is either a typing error or a grade written against a different standard.
• h — a harder base binder. The trailing h changes exactly one requirement: the penetration of the distillation residue at 25 °C. CSS-1 residue must fall between 100 and 250 dmm; CSS-1h residue must fall between 40 and 90 dmm. Viscosity band, residue minimum, cement mixing limit, sieve limit, particle charge, ductility and solubility are identical. The h grade is discussed in full further down this page, because getting it wrong is the most common specification error on a slow-setting order.

Breaking and curing are two different events

Almost every misunderstanding about slow-setting emulsion comes from treating these as one thing. Breaking is the moment the emulsion stops being an emulsion: the droplets coalesce, the water separates out as a distinct phase, and the film changes colour from brown to black. Curing is what happens afterwards, as the separated water leaves the site by evaporation, by draining away or by being absorbed into the layer beneath, and the binder film gains its final strength and adhesion.

For a cationic emulsion the break is driven mainly by chemistry rather than by weather. The amine head of the emulsifier is adsorbed onto the negatively charged mineral surface, the charge holding the droplets apart is stripped away, and the bitumen deposits onto the stone. That is why cationic grades break more predictably than anionic ones, which on siliceous stone are left waiting for the water to evaporate. A slow-setting grade slows the first event by carrying enough emulsifier that the adsorption does not immediately destabilise the remaining droplets. It does not, and cannot, slow the second event indefinitely: once broken, a thin CSS-1 film cures quickly because there is very little water in it.

The practical consequence: a grade that can be worked cold

Because CSS-1 has a Saybolt Furol viscosity of only 20 to 100 seconds and that measurement is taken at 25 °C, it is a genuinely cold-applied material. It pumps and sprays at ambient temperature in most climates, without a heated distributor, without a hot oil jacket and without the thermal precautions that a paving grade or a heavy rapid-setting emulsion demands. CRS-2, by contrast, is specified at 100 to 400 SFS at 50 °C and is normally handled warm. The difference in test temperature on the certificate is not a laboratory convention; it reflects the temperature at which each grade is actually used, and comparing a 25 °C figure against a 50 °C figure is one of the more damaging misreadings of an emulsion certificate.

What CSS-1 is not

• Not a spray seal binder. A chip seal needs a thick binder film that breaks under the chippings before the roller arrives. A slow-setting grade under chippings is still brown when the roller reaches it, and the stone sheds under the first traffic.
• Not the same as SS-1. SS-1 is the anionic slow-setting grade under ASTM D977: negatively charged droplets, alkaline product, different acceptance limits on the residue penetration band. The two must never share a tank, a line or a pump, because combining the charge families neutralises both and coagulates the bitumen immediately.
• Not automatically CSS-1h. Suppliers, engineers and purchase orders drop the h far more often than they should. It is a different residue and, in a hot climate, a different outcome at the layer interface.
• Not a cutback. There is no solvent in a CSS grade, ASTM D2397 sets no oil distillate requirement for the slow-setting family, and there is therefore no flash point requirement and none of the Class 3 fire case that governs MC-30 or any other cutback. It brings a completely different, and much smaller, hazard set.
• Not a diluted CRS-2. Adding water to a rapid-setting emulsion does not create a slow-setting one. It creates an unstable rapid-setting emulsion with a thin film, which is a description of a failed seal.

Put in one sentence: a slow-setting emulsion stays fluid long enough to be mixed with fine aggregate, or to be diluted and sprayed thinly and evenly, which is exactly what a tack coat, a slurry seal or a soil stabilisation job needs and what a rapid-setting grade cannot do. Everything below expands that sentence.

Surface area is the enemy, and fines have a great deal of it

An emulsion breaks when droplets meet a mineral surface and their stabilising charge is stripped away. The rate at which that happens depends on how much mineral surface the emulsion is asked to meet. A single-size chipping in a surface dressing presents a modest amount of surface per kilogram of stone. A well-graded sand, a crusher dust, a soil containing silt and clay, or Portland cement presents orders of magnitude more surface per kilogram, and every square metre of it is an opportunity for the emulsion to break in the wrong place at the wrong moment.

That is why the setting class, not the charge and not the viscosity, decides whether a grade may be put in a mixer. A rapid-setting grade in a pugmill breaks in the pugmill. The mix seizes, the machine has to be cleaned out, and the day’s production is lost. A medium-setting grade survives coarse open-graded aggregate but not fines. Only a slow-setting grade survives fines, and ASTM D2397 makes that a testable claim rather than a marketing one.

The cement mixing test is the proof, and it is the hardest thing you can ask of an emulsion

The class-defining acceptance test for CSS-1 and CSS-1h is cement mixing to ASTM D6935: the emulsion is mixed with high surface area Portland cement, and the coagulum retained on the sieve afterwards must not exceed 2.0 %. Cement is chosen because it is the most aggressive fine material the grade will ever meet in service — enormous specific surface, and chemically active in water on top of it. A grade that survives cement will survive a graded sand, a crusher dust or a silty subgrade soil.

Read that test as the buyer’s guarantee that the emulsion can be mixed at all. A CSS-1h that fails cement mixing will flash-set in a slurry seal machine, and the first sign will be a box that will not discharge. It is worth confirming the result is present on the certificate before a slurry or stabilisation contract starts, because it is the line that a rapid-setting grade mislabelled as slow-setting cannot fake.

Dilution is only available to this class

The second thing a slow break buys is the ability to add water. Dilution reduces the concentration of emulsifier in the water phase as well as thinning the product, so a grade that is only just stable neat becomes unstable when it is cut. Rapid-setting grades are therefore never diluted — and in a chip seal there would be no reason to, because thinning the binder destroys the film thickness the chippings need. Medium-setting grades are not diluted either; their water content is set by the cold-mix design.

Slow-setting grades are the exception, and dilution is not a way of stretching the drum. It is a spraying technique. A distributor cannot lay a fifth of a litre per square metre of neat emulsion evenly: the pump has a minimum controllable output, the nozzles have a minimum flow before the fan collapses, and the result of trying is a striped, ropey pattern with bare bands between the fans. Cutting the same residual binder into two or three times the volume of liquid moves the sprayed rate up into the range the bar can actually control, and the film that lands is thin, continuous and even. That is the whole argument for dilution, and it is why the section below is the longest on this page.

Thin and even is what a bond coat needs

A tack coat — a bond coat, in the language of most European specifications — is not a structural layer. It is an interface treatment whose job is to make two asphalt layers behave as one. The quantity of binder involved is small, the tolerance on it is tight in both directions, and coverage matters more than thickness. A thin continuous film bonds; a thick film lubricates; a broken film leaves unbonded areas that become the starting point for delamination. Delivering a small quantity of binder evenly across a whole carriageway width is a much harder problem than delivering a large quantity, and it is a problem that only a thin, dilutable, cold-applied, slow-breaking liquid solves. The full comparison of what a bond coat does against what a prime coat does is on prime coat vs tack coat.

The jobs that exist because slow-setting grades exist

• Tack coat between pavement layers. The dominant use of CSS-1 and CSS-1h by volume. Diluted, sprayed thin and allowed to break and cure before the paver arrives.
• Fog seal. A light, heavily diluted spray applied over an existing surface to seal fine cracks, arrest ravelling and enrich a dry, oxidised wearing course, often with a light sand blotter so the road can be reopened. The dilution here is heavier than for tack work precisely because the quantity of binder wanted is smaller still.
• Slurry seal. Emulsion, fine graded aggregate, mineral filler and water mixed in a continuous machine and spread as a thin surfacing. This is impossible with anything but a slow-setting grade, and CSS-1h is the usual choice because the residue carries traffic directly. Slurry systems are specified through ISSA recommended performance guidelines — ISSA A105 for slurry seal — rather than through ASTM D2397, which specifies the emulsion and not the system.
• Soil and granular base stabilisation. The emulsion is mixed into a soil or granular material to bind the fines and reduce its sensitivity to water. The material being treated is largely fines, so the emulsion must tolerate them.
• Cold in-place recycling and full-depth reclamation. Milled pavement, or pavement plus base, mixed with emulsion and relaid. The mixing window has to be long enough to place and compact the material, which again rules out a rapid break.
• Dust control on unpaved surfaces. Heavily diluted and applied to haul routes, shoulders and unsurfaced access roads to bind surface fines. The most water-tolerant application of the lot.
• Priming tight granular bases where a cutback is restricted. A slow-setting emulsion will wet and bind a tight, fine-graded base surface. Be precise here: a plain CSS-1 is not a purpose-made penetrating emulsion prime, and specialist prime emulsions exist that are formulated to penetrate rather than to sit. Where the base is open and stony, a cutback still does something an emulsion struggles to do. The comparison sits on the prime coat and tack coat page.

What the slow break takes out of the programme

The trade is real and it belongs in the programme rather than in a footnote. A rapid-setting chip seal can be swept and opened within the working day. A slow-setting tack coat has to break and then cure before anything runs over it, and until it does, it tracks: construction plant picks the binder up on its tyres, carries it down the road and leaves bare stripes in the wheel paths where the bond was supposed to be. Nothing about the chemistry can be hurried. The break window depends on the dilution, the surface, the temperature, the humidity and the wind, which is why it is set by observation on the day rather than by a number on a data sheet. Planning around it — spraying only as far ahead as the paver will reach, and keeping every other vehicle off — is the single most useful thing a site can do with the information on this page.

The only line that changes

Under ASTM D2397 the residue recovered by distillation must be penetration tested at 25 °C to ASTM D5. For CSS-1 the result must fall between 100 and 250 dmm. For CSS-1h it must fall between 40 and 90 dmm. Every other requirement is word-for-word identical: Saybolt Furol viscosity 20 to 100 seconds at 25 °C, residue by distillation minimum 57 % by mass, cement mixing maximum 2.0 %, sieve maximum 0.10 %, particle charge positive, 24-hour storage stability maximum 1 %, five-day settlement maximum 5 %, residue ductility minimum 40 cm and residue solubility minimum 97.5 %.

The plain-language version is that CSS-1h is the same emulsion made from a harder base binder. The emulsifier system, the manufacturing route through the colloid mill, the water content, the handling rules, the dilution practice and the storage window are unchanged. What changes is the bitumen charged to the mill, and therefore the stiffness of what is left on the road after the water has gone.

Why a harder residue is worth specifying

The residue is the binder the pavement keeps. At the interface between two asphalt layers it sits in a very thin film, and in summer it sits there hot. A pavement surface in a hot inland climate reaches temperatures at which a 100 to 250 dmm binder is extremely soft, and the interface film is being asked to transfer shear between two layers while a heavy, slow-moving or braking wheel passes over it. A softer residue at that interface has less shear capacity, and the failure it permits is slippage — the crescent-shaped cracking described in the failure table below.

That is why the h grades exist and why they dominate in hot climates and on heavily trafficked roads. Three situations point clearly at CSS-1h:

• High summer pavement temperatures. Where the wearing course is specified with a hot-climate binder, the bond coat under it should not be the softest thing in the structure.
• Heavy, slow or channelised traffic. Climbing lanes, bus lanes, roundabout approaches, port and terminal roads, signalised junctions — anywhere the wheel loads are high and the horizontal force at the interface is high with them.
• Slurry seal. In a slurry the residue is not an interface film at all; it is the binder of the surfacing, carrying traffic directly. CSS-1h is the normal slurry grade for that reason.

Conversely, CSS-1 with its softer residue is a reasonable choice for fog seals on an oxidised surface, for dust control, for soil and granular base stabilisation, and for tack coats in cooler climates or under light traffic, where the softer binder wets and penetrates a little more readily and the shear demand at the interface is modest.

The residue penetration is not the base binder penetration

A point that causes arguments at the certificate stage. The residue tested under ASTM D5 is the material recovered by distillation to 260 °C under ASTM D6997, and that distillation is itself a thermal ageing step. The residue is therefore harder than the bitumen that was charged to the colloid mill, and a residue penetration of, say, 70 dmm does not mean the manufacturer used a 70 dmm base binder. The specification is written on the residue, the acceptance decision is made on the residue, and no useful inference about the feedstock runs backwards from it. If your specification is genuinely about the base binder — which is unusual — that has to be written as a separate requirement and agreed, because it is not what ASTM D2397 measures. Background on the base grades themselves is on the penetration grade bitumen page.

How the h gets lost, and how to stop it

The failure mode is administrative rather than technical, and it is common. An engineer writes CSS-1h into the specification. The bill of quantities item says cationic slow setting emulsion. The purchase order says CSS-1. The supplier, entirely correctly, ships CSS-1 and certifies a residue penetration of 140 dmm, which conforms. Nobody notices until a hot summer and a heavy vehicle produce slippage cracking at a junction, at which point there is a conforming certificate for the wrong product.

Three habits prevent it:

• Name the grade in full on every document, including the h, from the specification through the enquiry and the purchase order to the certificate. Slow setting cationic emulsion is not a grade name.
• Check the residue penetration line on the certificate against the grade named on the order. A result between 100 and 250 dmm is CSS-1; a result between 40 and 90 dmm is CSS-1h. That single line settles which product arrived, and it is the only line that does.
• Ask which standard the grade is written against. The C prefix is an ASTM convention, not a universal one. Under the Indian standard IS 8887, which specifies cationic bitumen emulsions for roads, the slow-setting grades are written SS-1 and SS-2 with no C prefix and no h suffix, because every grade in that standard is cationic by the standard’s own scope. An SS-1 on a document written to IS 8887 and an SS-1 on an ASTM D977 certificate are opposite charges. Under EN 13808 the grammar is different again: a cationic emulsion is written in the form C60B4, with the number giving binder content by mass, B indicating a straight bitumen binder and the final digit giving a breaking-behaviour class determined by EN 13075-1. None of these systems maps one-to-one onto ASTM designations, so where a project names a European or Indian grade, buy against that standard rather than offering the nearest ASTM equivalent.

What the h does not change

It is worth being explicit, because the h is sometimes read as a general upgrade. It is not a stronger emulsion, it does not store better, it does not resist freezing, it is not more concentrated — the residue minimum is 57 % for both — and it does not change the dilution practice, the water quality requirement, the metallurgy, the sieve limit or the cement mixing limit. Everything in the dilution and handling sections below applies identically to CSS-1 and CSS-1h.

Why dilution happens at all

Start with the point that is most often assumed and is wrong: dilution is not a way of making the drum go further. It cannot be, because the binder in the drum does not increase when water is added. If the same residual binder rate is specified before and after dilution, the diluted material simply has to be sprayed at a proportionally higher rate, and exactly the same quantity of drums empties over exactly the same area.

Dilution is a spraying technique. The quantity of binder a bond coat needs is small, and a distributor cannot lay a small quantity of neat emulsion evenly. Below a certain flow the pump loses control of the rate, the pressure at the bar drops out of the range the nozzles were designed for, the fans collapse into ropes and the pattern lands as alternating heavy and bare longitudinal bands. Cutting the emulsion into two or three times the volume of liquid raises the sprayed rate back into the band where the bar behaves properly, and the film that lands is thin, continuous and even. That, and the fact that a thinner liquid wets a dry or dusty surface a little more readily, is the whole justification.

The specification is in residual binder, and that is the number to work from

A properly written tack coat specification states a residual binder rate — kilograms or litres of binder per square metre after the water has gone — not a rate of emulsion sprayed. It has to, because a rate written as litres of emulsion is undefined until somebody also states the residue content and the dilution, and those two variables can move the delivered binder by a factor of four.

Where a specification is written the other way round, in litres of emulsion per square metre, the first job on site is to convert it into residual binder and then reconstruct the sprayed rate for the material actually in the tank. If the specification does not say what dilution it assumed, ask the engineer. Do not guess, and do not assume neat.

The arithmetic, in three steps

Two assumptions make this tractable and both should be stated openly. First, the density of a bitumen emulsion is close to 1.0 kg per litre, so litres and kilograms are interchangeable to within the accuracy of a distributor. Second, dilution water also has a density close to 1.0 kg per litre, so a dilution stated by volume and a dilution stated by mass come to the same thing. Both are approximations, both are good enough for a spray rate, and both should be replaced by the supplier’s actual density figure where a contract is being settled on quantity.

1. Take the residue content of the material you actually have. Use the measured residue on the batch certificate, not the ASTM D2397 minimum of 57 %. A batch measuring 60 % delivers about 5 % more binder at the same sprayed rate than a batch at the 57 % minimum, and on a long job that is not noise.
2. Work out the residue content of the diluted material. For a dilution of one part emulsion to n parts water by volume, the residue content of the mixture is the neat residue divided by (1 + n). At the 57 % minimum: 1:1 gives 28.5 %, 1:2 gives 19.0 %, 1:3 gives 14.3 %. A 2:1 dilution — two parts emulsion to one part water — gives 57 % multiplied by two thirds, which is 38.0 %.
3. Divide the specified residual rate by that figure. A specified residual binder rate of 0.15 kg/m² from a 1:1 dilution at 28.5 % residue needs 0.15 divided by 0.285, which is 0.53 litres per square metre of diluted material. The same 0.15 kg/m² from the neat 57 % product needs 0.26 L/m², which is a rate a distributor will struggle to lay evenly — which is the argument for diluting, made in numbers.

The mistake this arithmetic prevents

The classic failure runs like this. A rate of 0.53 L/m² is established for the 1:1 diluted material and written into the method statement. The next load arrives, dilution is skipped because the tank is already part full of neat product or because nobody was told, and 0.53 L/m² of neat CSS-1 goes down. That is 0.53 multiplied by 0.57, or 0.30 kg/m² of residual binder — twice what the specification asked for. The result is a lubricated interface rather than a bonded one, and the pavement above it shoves and cracks in crescents where vehicles brake and turn.

The reverse error is just as common and its consequences are slower to appear. A rate established for neat product is used on diluted material, delivering half the specified binder or less, and the interface debonds one, two or five years later — long after anybody is looking for a cause. Write down the residual rate, the residue content used, the dilution used and the resulting sprayed rate together, as four numbers on the same line, and re-derive all four whenever any one of them changes.

How to dilute without breaking the emulsion

The mechanics matter as much as the ratio, and the rules are short:

• Add the water to the emulsion, slowly, with gentle agitation. Never pour emulsion into a tank of water. An emulsion introduced into a large volume of water is momentarily surrounded by far more water than its emulsifier can stabilise, and it breaks on the spot.
• Match the temperatures. A large temperature difference between the emulsion and the dilution water is a shock to the system and can break it. Bring the two within a few degrees of one another before mixing.
• Agitate gently and briefly. Enough to blend, not enough to shear. High shear pumping and violent circulation coalesce droplets.
• Dilute only what the day will use. Diluted emulsion is less stable than neat emulsion, and it settles faster because the same bitumen droplets are now falling through a thinner, less viscous liquid. Leaving diluted material to stand overnight in a distributor is asking for a separated tank and a blocked bar in the morning.
• Sieve test the diluted material before committing a day to it. A quick ASTM D6933 sieve check on the diluted product is the quickest and simplest confirmation that the water has not broken it. It is particularly worth doing the first time a new water source is used.

The water is a raw material

This deserves its own heading because it is genuinely the thing that breaks emulsion in the tank, and it is treated as a utility on most sites. The rules are set out in full in the handling section below, but the headline is simple: a cationic emulsion is stabilised by an amine emulsifier that only carries its positive charge while the product stays acidic, so alkaline water strips the charge and breaks the emulsion. Hard, alkaline, silty, saline or contaminated water can coagulate a tank of CSS-1 without anyone touching the pump. Clean potable water, and a laboratory dilution trial before a new source is used at scale.

Where dilution ratios come from

ASTM D2397 sets no dilution ratio, because it specifies the product and not the operation. The ratios in general use are industry practice and project specification, and they vary widely by country, by agency and by job:

• Tack coat is very commonly diluted 1:1 with clean potable water, which is the figure most agency guidance and most manufacturer literature converges on.
• Fog seals are normally diluted more heavily than tack coats, because the binder quantity wanted is smaller and the film wanted is thinner still.
• Dust control applications use the heaviest dilutions of all.
• Slurry seal and micro-surfacing use the emulsion neat; the water in the mix is part of the mix design and is added separately at the machine, not to the emulsion.
• Soil and base stabilisation dilutions are set by the mix design and by the moisture condition of the material being treated.

Treat every one of those as a starting point for a trial, confirm the ratio against the project specification and the supplier’s data sheet for the grade actually delivered, and then convert it into a sprayed rate using the measured residue on the batch certificate. Where the specification and the data sheet disagree, the specification governs the acceptance and the data sheet governs the product.

The rule that admits no exception: it must never freeze

The continuous phase is water. Below 0 °C ice crystals grow through the emulsion and mechanically rupture the emulsifier films around the droplets; the droplets merge, and what thaws is a layer of coagulated bitumen under a layer of dirty water. The change is irreversible. No amount of heating, stirring or recirculation restores it, and no grade in either ASTM specification is exempt — CSS-1 and CSS-1h included.

The practical floor is not 0 °C but about 4 °C, and the reason is measurement rather than chemistry. The coldest part of a consignment is never the part anybody measures: the windward drum in a stack, the top of a part-filled tank, the outer container in a stow and the drum standing on cold concrete all run below the bulk. A four-degree margin covers the gap between the thermometer and the worst-off drum. This is industry practice and manufacturer guidance rather than a requirement of ASTM D2397, which sets no storage temperature at all — but it is practice with no dissenting view behind it.

Because the rule is absolute, it belongs in the shipping decision rather than in the site method statement. There is no cold-climate CSS grade to switch to. Insulated or heated storage at a winter destination is arranged before the cargo sails, the routing is checked for cold-weather transhipment, and samples travelling to a laboratory are protected in transit as carefully as the cargo is.

The upper limit is lower than a CRS-2 tank, and the grade name tells you so

Above roughly 85 °C — a figure carried in manufacturer literature and general emulsion practice, not a limit set by ASTM D2397, which specifies no temperature of any kind — water flashes off the surface of any emulsion, a skin forms, the product thickens and eventually breaks. But the working window is grade-specific, and the slow-setting grades sit well below the heavy rapid and medium grades. Typical recommended windows — again manufacturer guidance and common practice, not ASTM requirements — put CSS-1, CSS-1h, SS-1, SS-1h and MS-2 in the region of 10 to 60 °C, against about 50 to 85 °C for CRS-1, CRS-2, RS-2, CMS-2 and CMS-2h.

The practical trap is a shared yard. A tank thermostat left where a CRS-2 consignment needed it will skin and thicken a CSS-1, and the first symptom will be a viscosity result out of the 20 to 100 SFS band and a sieve result above 0.10 %. Check the label before setting the thermostat, and confirm the window against the supplier’s data sheet for the grade actually delivered; where the two disagree, the data sheet governs. Tank design, coil surface loading and instrumentation are covered on the bitumen storage tanks page.

Circulation: gentle, submerged, and only when it is needed

Bitumen droplets are slightly denser than water and drift downward given time, which is what the 24-hour storage stability and five-day settlement tests to ASTM D6930 are measuring. Circulation exists to reverse that drift, and it is genuinely useful — a settled but unbroken consignment usually comes back with gentle recirculation. Done badly, it is one of the more reliable ways to break a tank.

• Low shear, low speed. High-shear pumping coalesces droplets by forcing them together mechanically. Positive displacement pumps run at low speed are the normal choice; a centrifugal pump run hard through a restricted line is not.
• Keep the return line submerged. A return that discharges above the liquid surface entrains air, whips the surface and builds a skin. It should discharge below the surface, and preferably near the bottom.
• Circulate briefly and occasionally, not continuously. Continuous circulation is not conservative practice; it is prolonged mechanical work on a product that is held together by nothing but electrostatic repulsion.
• Do not circulate through the spray bar. The bar is a set of small orifices at pressure, which is a high-shear device by definition. Circulate through the return line.
• Avoid free fall on filling. Fill from the bottom or through a dip pipe reaching below the liquid surface. A stream falling through air aerates the product and skins it.
• Never pump emulsion into a hot vessel. Water flashing to steam expands more than a thousandfold. Emulsion of any designation pumped into a tank, line or tanker still above 100 °C, or onto a residue of hot binder, can eject the contents through the hatch. Confirm every receiving vessel is cool, drained and free of hot bitumen before transfer, and never steam-clean a line into a live emulsion tank.

Dilution water is a raw material, not a utility

This matters more for CSS-1 than for any other property on the page, because dilution is the normal way this grade is used and because the water is the one component nobody puts on a certificate.

A cationic emulsion is stabilised by a fatty amine emulsifier that has been neutralised with acid. It carries its positive charge only while the product stays acidic — the cationic family is typically in the region of pH 2 to 4, which is a characteristic of the family rather than a limit either ASTM specification sets. Raise the pH and the amine is deprotonated, the charge that holds the droplets apart disappears, and the emulsion breaks. That is the mechanism, and it explains why the water matters:

• Alkaline water is the direct hazard. Water with high bicarbonate alkalinity neutralises the acid that is keeping the emulsifier charged. Groundwater in limestone country is the classic offender, and so is water that has stood in a concrete tank.
• Hard water is usually alkaline water. Hardness and alkalinity travel together in most natural supplies, which is why the practical advice is expressed as clean potable water of low hardness. The dissolved calcium and magnesium bring their own destabilising effects as well.
• Suspended solids break emulsion the same way aggregate does. Silt and clay in the water are fine mineral surfaces with a great deal of area, delivered directly into the tank. River water, canal water and water pumped from a site sump are all suspect.
• Saline water and seawater are out. Dissolved salts destabilise the system, and the chlorides attack the equipment as well.
• Recycled, waste and process water are out, and so is water from a tank that has held detergent, soap, solvent or an anionic product. Anionic surfactant residues neutralise a cationic emulsion on contact.

The practical protocol is short. Use clean potable water. Where the source is new, or where the site supply is a borehole rather than a mains connection, run a small laboratory dilution trial at the intended ratio before committing the job to it: mix a sample of the actual emulsion with the actual water, let it stand, and check it on an 850 µm sieve to ASTM D6933. It takes about an hour and it is the only reliable way to find out that a water source is unusable before there are ten thousand litres of coagulum in a distributor.

Metallurgy and cross-contamination

Because the cationic family is acidic, everything the emulsion touches must be mild steel or stainless steel: tank, lines, pump, valves, couplings and bar. Aluminium, zinc, galvanised coatings and copper alloys corrode, and the dissolved metal ions destabilise the emulsion as they go — so the first symptom is often a failed sieve test rather than a visible leak. A site commissioning second-hand tankage for its first emulsion consignment should audit the wetted materials before the cargo arrives.

Cross-contamination is the other half of the same discipline. A cationic and an anionic emulsion combined — even by pumping one through a line still wet with the other — neutralise both charges and coagulate the bitumen immediately. Never share a tank, a line, a pump or a distributor between the two families without a documented flush, and use the particle charge test to ASTM D244 to settle any doubt about what a vessel last held.

Safety: a much smaller hazard set than a cutback, but not an empty one

• No solvent, no flash point requirement. ASTM D2397 sets no oil distillate limit for the CSS grades, they are supplied without a solvent fraction, and there is no flash point line on the specification. None of the Class 3 fire case that governs a cutback applies here. Cationic bitumen emulsion is not normally shipped as dangerous goods — but confirm the classification from the supplier’s Safety Data Sheet for the product actually supplied rather than assuming it, because that is the document a carrier and a customs authority will read.
• Do not generalise that to the emulsion family. The absence of a solvent is a property of the CSS grades, not of emulsions. ASTM D2397 permits the rapid-setting CRS grades up to 3 % petroleum distillate by volume and the medium-setting CMS grades up to 12 %, and that distillate partitions into the vapour space of a warm drum or tank, so the vapour above a warm CRS or CMS drum can be flammable. A yard that handles more than one emulsion grade must read the label and the Safety Data Sheet before it decides what the fire case is.
• Never heat an emulsion with an open flame or a bare electric element, and keep ignition sources away from drums and tanks. Indirect, low-intensity heating only — submerged coils, hot oil or steam. This is not merely a quality rule. A burner or bare element against a drum or tank wall boils the water phase at the metal, breaks the emulsion onto the heating surface and can pressurise a closed vessel through the bung; and in a mixed yard the same habit applied to a CRS or CMS drum, which does carry a distillate fraction, puts a naked flame beside a flammable vapour space. Bottle burners, torches, brazier fires and open kettles have no place near any emulsion drum or tank. No smoking, no hot work, no welding, cutting or grinding beside stored emulsion.
• The product is acidic. Eye and skin contact cause irritation. Goggles or a face shield when decanting or connecting hoses, nitrile gloves, and immediate flushing with clean water on contact.
• Steam is the real physical hazard. Every point about hot vessels above applies to drums and road tankers as much as to storage tanks: water flashing to steam expands more than a thousandfold, and a vessel, line or tanker still above 100 °C, or holding a residue of hot binder, can eject its contents through the hatch when emulsion is pumped in.
• Hot bitumen and emulsion never share equipment. Not warm binder run onto emulsion, not emulsion pumped onto the warm residue of a paving grade, and not a shared transfer line.

Packing, and why paving-grade arithmetic does not carry across

Emulsion is normally traded by volume rather than by mass, and the commercially meaningful quantity is neither: it is the residual binder. Both points make a difference to how an enquiry should be written.

Take the container arithmetic honestly. The site standard for a 20' FCL of drummed product is 80 drums, and for paving grades that is quoted as 12 MT at a 150 kg fill, 14.4 MT at 180 kg or 14.8 MT at 185 kg. Those are mass fills of a hot-poured product, and they do not transfer to emulsion. An emulsion is filled cold, by volume, into a new 210 litre drum with ullage left for expansion, and its density is close to 1.0 kg per litre, so the net weight follows the fill volume rather than being set as a nominal mass. On the fill this site quotes for drummed emulsion — about 180 kg net, which is roughly 180 litres in a 210 litre drum — eighty drums come to about 14,400 litres and about 14.4 MT of emulsion. At the ASTM D2397 minimum residue of 57 % by mass, that container carries roughly 8.2 MT of binder. Set that against a paving-grade container at the same 180 kg fill: 14.4 MT declared, and 14.4 MT of binder, because the drum contains nothing else. The tonnage on the bill of lading is the same and the binder inside it is little more than half, which is exactly why quoting emulsion by tonnage alone tells a buyer very little. Note also that the container figure moves with the fill: at the heavier 185 kg fill the same eighty drums are about 14.8 MT of emulsion and about 8.4 MT of binder, so the net fill per drum belongs on the packing list rather than being assumed.

Two corollaries follow. First, the jumbo and meltable bag arithmetic used elsewhere on this site — 1 MT bags, 20 bags and 20 MT per 20' FCL — belongs to grades that are solid at ambient temperature and has no application whatever to a liquid emulsion, which cannot be bagged at all. Second, an emulsion enquiry should state the fill volume per drum, the number of drums, the total litres and the expected residue content, and offers should be compared on residual binder rather than on emulsion tonnage. The conversion background is on tonnage and volume conversions, and the packing options themselves on bitumen packaging.

• New steel drums with sound closures, correctly labelled with the grade in full including any h suffix and with the manufacturing date. Reconditioned drums are a contamination risk for emulsion in a way they are not for a paving grade, because a residue of the wrong charge family destroys the contents. See bitumen in new steel drums.
• Drums upright, bungs tight, off cold ground and out of direct sun, shaded and ventilated. Do not stack beyond the drum rating, and do not let a stack sit against an unheated wall in a winter destination.
• Protect from freezing at every point in the chain, including the load port apron, the transhipment stow and the destination yard, not only in the final store.
• Ask for the manufacturing date and batch number on the certificate. Emulsion is the one bitumen product family whose properties move measurably while it sits, so an undated certificate cannot be interpreted. The shelf life question is treated on bitumen shelf life and storage, and sampling on bitumen sampling procedure.